基于CTA影像组学特征的前交通动脉瘤破裂的预测模型的构建及验证

Construction and validation of a prediction model for rupture of anterior communicating artery aneurysms based on morphological characteristics of CTA radiomics

目的探讨基于CTA影像组学特征构建预测前交通动脉瘤破裂的模型的临床价值。方法回顾性分析2016~2023年收治的116例前交通动脉瘤病人的病历资料,按照8:2的比例分配为训练集和测试集。收集入院头部CTA影像参数,应用3D Slicer软件提取影像组学特征,并计算影像组学评分(Rad score);多因素logistic回归模型分析动脉瘤破裂的危险因素并构建预测模型,应用ROC曲线评估模型的预测能力,并应用决策曲线评估模型的临床应用价值。结果多因素logistic回归分析显示年龄(OR=0.944;95%CI 0.897~0.993;P=0.025)、瘤体长度与载瘤动脉管径的比值(SR;OR=2.247;95%CI 1.214~4.15;P=0.016)和瘤高与颈宽的比值(AR;OR=7.942;95%CI 1.47~42.925;P=0.010)为动脉瘤破裂的独立预测因素。Lasso回归模型筛选出有意义的4个影像组学特征(Maximum 2D DiameteRColumn,Maximum 2D DiameteRRow,Surface Volume Ratio,Elongation),经计算获得Rad score。ROC曲线分析显示,基于年龄、AR、SR及Rad score构建的联合模型,训练集曲线下面积(AUC)为0.889(95%CI 0.821~0.958),测试集为0.921(95%CI 0.803~0.999)。校准图显示实际概率与预测概率之间具有良好的预测精度,决策曲线显示在阈值概率37%~65%范围内,联合模型净收益高于传统影像预测模型。结论CTA影像组学特征联合传统影像学特征构建的联合模型对前交通动脉瘤破裂有较好的预测能力。

Objective To explore the clinical value of constructing a model for predicting the rupture of anterior communicating artery aneurysms based on CT angiography(CTA)radiomics features.Methods The medical records of 116 patients with anterior communicating artery aneurysms admitted from 2016 to 2023 were retrospectively analyzed and divided into a training set and a test set at a ratio of 8:2.Head CTA imaging parameters were collected,radiomics features were extracted using 3D Slicer software,and the radiomics score(Rad score)was calculated.A multivariate logistic regression model was used to analyze the risk factors for aneurysm rupture and construct a prediction model.The predictive ability of the model was evaluated using the ROC curve,and the clinical application value was evaluated using the decision curve.Results The multivariate logistic regression analysis showed that age(OR=0.944;95%CI 0.897~0.993;P=0.025),the ratio of aneurysm length to parent artery diameter (SR;OR=2.247;95%CI 1.214~4.15;P=0.016),and the ratio of aneurysm height to neck width(aspect ratio,AR;OR=7.942;95%CI 1.47~42.925;P=0.010)were independent predictors of aneurysm rupture.FouRsignificant radiomics features(Maximum 2D Diameter Column,Maximum 2D Diameter Row,Surface Volume Ratio,Elongation)were screened out by the Lasso regression model,and the Rad score was obtained through calculation.The ROC curve analysis showed that the area under the curve(AUC)of the combined model based on age,AR,SR,and Rad score was 0.889(95%CI 0.821~0.958)in the training set and 0.921(95%CI 0.803~0.999)in the test set.The calibration plot showed good predictive accuracy between the actual probability and the predicted probability.The decision curve showed that within the threshold probability range of 37%~65%,the net benefit of the combined model was higher than that of the traditional imaging prediction model.Conclusion The combined model constructed by combining CTA radiomics features with traditional imaging features has a good predictive ability for the rupture of anterior communicating artery aneurysms.